Overview
**Happy Horse Edit 1080p** is a video model for generating and editing short HD clips from prompts, source videos, and reference images. The public model id is `happy-horse-edit-1080p`.
It is designed for workflows that need controlled video transformation rather than long-form rendering. The model supports asynchronous generation and can process clips up to the first 15 seconds, making it suitable for fast creative iteration, product visuals, ads, social media assets, and automated video pipelines.
What the model supports
Happy Horse Edit 1080p supports three main generation modes:
- **Video edit**: edit an uploaded source video using a text prompt.
- **Image to video**: generate motion from one or more reference images.
- **Reference media to video**: create a video guided by prompt and media references.
Key parameters include:
- `prompt`: required text instruction.
- `duration`: 3 to 15 seconds.
- `resolution`: `720p` or `1080p`.
- `video_url`: source video for video editing.
- `reference_image_urls`: 1 to 5 reference images for image-guided generation.
- `video_count`: fixed at 1 video per generation.
Pricing
For the 1080p variant, UniAll AI lists Happy Horse Edit 1080p at **$0.3808 per second** for user pricing. Billing is based on video duration.
Example estimated costs:
| Duration | 1080p estimated cost | |---:|---:| | 3 seconds | $1.1424 | | 5 seconds | $1.9040 | | 10 seconds | $3.8080 | | 15 seconds | $5.7120 |
A lower-cost 720p variant is also available at **$0.1904 per second**. Choose 1080p when the output needs sharper detail for ads, product showcases, presentation assets, or post-production review. Choose 720p for drafts, testing, or high-volume experiments.
Failed generations are refunded when supported by the platform policy. Automatic retry is not enabled for this model, so production systems should handle failure states explicitly.
API usage
Endpoint:
```http POST /v1/videos/generations ```
Example request:
```json { "model": "happy-horse-edit-1080p", "generation_mode": "video_to_video", "prompt": "Enhance the product shot with cinematic lighting, smoother camera motion, and a premium studio look.", "duration": 5, "resolution": "1080p", "video_url": "https://example.com/source.mp4", "video_count": 1 } ```
For image-to-video workflows, use `generation_mode: "image_to_video"` and provide `reference_image_urls` instead of `video_url`.
Who should use it
Happy Horse Edit 1080p is a good fit for:
- Marketing teams producing short ad variations.
- E-commerce teams creating product display clips.
- Developers building AI video generation workflows.
- Agencies that need repeatable short-form video editing.
- Platforms that want API access to 1080p video editing without managing model infrastructure.
Practical guidance
Use precise prompts that describe the desired visual change, camera behavior, lighting, subject consistency, and output style. For video editing, keep source clips concise because the model processes up to 15 seconds. For cost control, test prompts at 720p or shorter durations before generating final 1080p outputs.
常见问题
How is Happy Horse Edit 1080p API priced?
Happy Horse Edit 1080p is billed per second. The listed UniAll AI user price for `happy-horse-edit-1080p` is $0.3808 per second for 1080p output.
What inputs does happy-horse-edit-1080p support?
The model supports text prompts, source video URLs for video editing, and 1 to 5 reference image URLs for image-to-video or reference-guided generation.
What is the maximum video duration?
The supported duration range is 3 to 15 seconds. For source video editing, the model processes up to the first 15 seconds.